From $0.61 Fixes to $300K in Sales: Adam Sand on AI That Works While You Sleep

Adam Sand, founder of Roofing Business Partner, shows how small AI automations generate outsized returns: a $0.61 workflow fix saved $1,500 monthly in context-switching costs, while a voice agent improved close rates from 28% to 30.8%, unlocking roughly $300K in additional sales.
Questions this video answers
August 28, 2025.
He said they ran $7,000 in ads and sold $400,000 worth of roofs.
$0.61 for the month.
Close rate increased from 28% to 30.8%, which Adam said translated to another $160,000 in sales.
369 digital leads per month; 61 of them had never booked an appointment; ~80% of those were evening/weekend leads.
Chapters
We ran seven grand with the ads and sold $400,000 worth of roofs.
So today's guest is one of the sharpest minds in the home service space.
So it's costing us $1,500 to $2,000 a month to have this problem. We built this automation. What do you think it would cost to fix it? The cost to do that was for the month was 61 cents.
And today's eaters share a story that just shows how much opportunity still exists in the trades.
Their close rate went from 28% to 30.8%, which translated to another $160,000 in sales.
If you think AI is going to kill jobs, stay tuned because conversation might change the way you look.
Technology, home services contractors have that problem.
This is AI and marketing for home service pros, the show for contractors who want to grow without wasting time or money. If you're in roofing, H vac, plumbing, or any home service business, we break down how AI and modern marketing can help you book more jobs, close more leads, and build a business that runs without babysitting every task.
All right, guys. So today's guest is one of the sharpest minds in the home service space, one of the few people that who could talk about AI philosophy and automation in the same sentence to actually make it work all for your business. I'm talking about Adam sand, founder of Roofing Business Partner, a company that's helped roofing actors scale using a mix of data driven systems, stoic principles, and AI power tools. He's a Christian, a father and a builder of both the roost and the people. And today he's here to share a story that just proves how much opportunity still exists in the trades.
Or.
Or if you think AI is going to kill jobs in the trades, Adam says the opposite. He believes it's going to create jobs and he's here to explain why. Stay tuned. Because conversation might change the way you look at commerce, technology growth, and the future of business. Let's get into it. Awesome.
Awesome.
Adam, welcome to the show.
Thank you very much, Bruce. It's great to be here. Thank you for making the time and space for me to come and chat with you. I'm excited. We always have good conversations.
Yeah. I first met Adam, I think seven years ago and he's one of the few people that was. Actually I learned how to do Facebook ads for roofing contractors. When I started the company, I was just doing PPC ads for a contractor. Directors.
Right.
For roofers. And eventually started doing Facebook ads and we added a call Center. But Adam was one of the few people a long time ago that had his own roofing business that was actually doing Facebook ads for himself, and he actually managed to teach other contractors to do the same as well.
Yeah, that was. That was. I remember we. The whole thing was we just. We were running Facebook ads in our roofing business, and everybody made fun of us for it. But I. I didn't know how to run Facebook ads out of the box. I had to take an online course. And even the teacher didn't really see a pathway to running Facebook ads for a home services business. Because back it was like weight loss and dog toys. But what happened was then when we did it, you know, in a world where you'd run a. Like, most of the people talking about Facebook ads back then would run like a small amount of money, like a thousand bucks, and they would sell like $4,000 worth of dog toys or like weight loss ebooks. So when we were like, oh, we ran seven grand with the ads and sold $400,000 worth of roofs, most of the market didn't know about net profit. They just saw clickbaity headlines. And so then all the. Everybody was trying to make their name in Facebook ads back then. And so we. We got interviewed on a bunch of podcasts. We, all of a sudden, like a year and a half later, if you looked up roofer and Facebook ad, my name showed up. And I was just talking about my results in my roofing business. And then I thought, wouldn't it be cool to start a podcast? Because all these roofers started reaching out to me so that maybe I should start a podcast. I was like, I didn't know what to call it. I was a partner in a roofing business, so I called a roofing business partner. That's as far as the name went. And then I just started doing it because I thought it would be kind of cool to have a podcast. And then I started helping roofers and then start this, like, school. And then it just. It was a. My whole. This all been one big acc. Accidental career. It was never intentional.
Yeah. Yeah. Awesome. So let's. Let's get started. So why should contractors start using AI?
Well, in some ways, because everybody else is going to, right? It's a certain degree is just. It's just you have to kind of get on the bed wagon. But then it comes down to, like, what is the real promise of AI? AI is really just another node on an automation. It's. There's this kind of joke in the community that's like when it's new, they call it artificial intelligence. And then when it's, when that becomes kind of mainstay and everybody understands, it becomes machine learning. And then after that becomes even more mainstay state, then they just call it automation, right? And so it's just the next node and you're in your automation in that. Now you have the ability to connect a tool, right? So your software and a memory to a large language model and give it context about your business so it can take action. So spoken financially, right? When you create, when you have a billion PhDs that just came on the labor market, AI and those billion PhD level brains are willing to work in the roofing industry, that's going to bring down the cost of labor. And it's going to, because they increase the supply of labor, like mental labor. And so if you start adopting a mindset around synthetic intelligence in your business, you can drastically increase the revenue per employee, which is going to make your business far more defensible. And I think that this is kind of table stakes at this point. And so thinking AI first is what's happening at the top level of business. And roofing is just one of those trickle down businesses that's the beneficiary of being old in a timeless vertical that's 100 years old and will live for another 100 years. We get the benefit of like seeing it play out in other industries and then just being there. But like anything else, the people who were there first tend to like gobble up the real estate or market share of that advantage. And I just want to like make that message clear so that roofers, and by extension the entire construction home services industry sees that there will be new market disruption options and new leverage created that can drastically change like any individual market share. So the small people don't have to be small and the big players might not necessarily stay big.
Yeah, exactly. What are some of the competitive advantages that you think a contractor, a tradesperson, a roofer can have by using AI? Like, like today.
All right, so let me break it down to like an example in my own business and then we'll kind of expand it out. So in my business we have this, we have this process where we, when every time we contractor sends us a work to do for them, a ticket, we open that ticket, we do the work and then we close the ticket and then it sends them a feedback survey and an email. So then sometimes they don't take the survey, they reply to the email and be like, thanks, great job. And Sometimes we call it 70, 30, 50, 50, right? Sometimes they're like, f you try harder, we got to like reopen the ticket. But every time they reply, our system opens the ticket. And our system only has a binary decision there. If you receive a reply to a closed ticket email, does it reopen the ticket or not? It's binary. It's the only decision it gets to make because that's all the intelligence it has. Did they rep apply or not? There's no context, right? Artificial intelligence allowed us to say, build a little workflow. And this workflow, no one's going to look at it in, in like na to N and be like, wow, so innovative. It's very, very simple. But this problem, right, of this ticket reopening meant that, okay, someone has now moved on to three other tickets and now they've got an email that says their ticket's been reopened. They don't know whether it's good, bad, what's happening. So then maybe they stop what they're doing. They context switch. So now they go, oh, I'm going to go back to this thing. And then they look at it and it was like, oh, they just said, than job. Now they go, ah, time for lunch. And then they come back and they have to completely re engage with that third ticket or they forget to go back to it. And then the customer's mad and then they have to go like it creates this context switching problem. And so this was conservatively costing us 1,500 to $2,000 a month, just in context switching extra hours, all this kind of stuff. So we built this very simple automation that just said when you get a reply, if it's a positive sentiment, like read it, if it sounds like, thank you, good job, appreciate you, leave the ticket closed if there's something else, we'll then summarize the content of the mess, reopen the ticket, assign a task that says like, this is what was happened. So then that way the person didn't get a notification about it, they just got a task which is like put itself onto the list. And I asked people, well, what do you think it cost me to put this synthetic brain in front of this? It was costing us 1500 to $2000 a month to have this problem. We built this automation. What do you think it would cost to fix it? Some people say 100, some people say a thousand. They're like, oh, probably cost you this much in automation cost or whatever, right? The cost to do that was for the month was 61 cents. So how many things in your business are costing you $1,000 or $1,500 a month or $500 a month. That could be fixed for 61 cents.
Yeah.
So what are some practical ways of doing that? It helps to know that there's something called like everybody hears the word agents and AI and there's some. There's a glossary of terms that I don't know if the listener has.
Yeah.
So we'll look at like a word like ambient agent. What is an ambient agent? So an ambient agent means it works in the background. It like an employee. This was that agent. It. It's not running all the time, it's waiting for something to happen. And then it takes an action and it's kind of ambient but it makes a decision on your behalf. And then you can choose three ways that an ambient agent works completely autonomously. Human in the loop where it says, hey, I need your help. I just need you to tell me what to do. This is what happened. What do I do? You click a button, say it sends a message to Slack and you have like a yes, no or more. Right. And you can just make one of three decisions and it goes. Or Human on the Loop where says, hey, this is what I did, but I think you should know about it. That's kind of what our $0.61 thing was. And then autonomous just means it's doing stuff and you have to like go look at it and figure out what happened. So if we think about it, well, now we kind of can start to think about what should be done. So a good example is a lot of roofing companies have an issue with like permits. Permits are something that they're spending someone to be around. You're still in Florida, right?
Yep.
So you have no commencements and permits and midday inspections. A huge logistics and bureaucracy problem in the Florida roofing market. Yes. So a roofing company has information in their CRM that informs that through the sales process this was collected and then it has to go to some person who sits at a desk all day and you know, sometimes it's a part of their job and if you get big enough, that is a person's whole job is just doing permits and notice of commencements. So what does that job entail? Well, you have to log into some old ass government website and fill out a bunch of boxes which is like, you know, name information, address, date of start, type of material we're putting on, expected install date, like just some data that exists in properties in the CRM. And previously we were never able to automate this because the government didn't say, hey, we released our zapier integration. So like what do you do? Well an ambient agent can use what's called a headless browser. So it's just like it sounds. It's a browser window that is headless. There's no person behind, just runs itself. And we can pre program that browser to understand this government website. And so it's like if the zip code is this, it belongs to this county and if it's this county, the website is this. And when you go to that website there will be these eight boxes to fill out and this one place where you need to upload this PDF. The eight boxes correlate to these eight properties in the CRM and the PDF is in this folder. Go fill out that form and then when you're done you're going to get a document back or you're going to have to save a, you're going to get a permit number or some kind of thing, scrape that out of the browser, copy paste, put it back into the CRM and then like check a box or move a stage in the pipeline view like then take an action. So that's an ambient agent. There's no human in loop, no human on the loop. It just happened. And that headless browser. Right. Allows you to say all right, well now we used to have to pay someone, you know, four hours a day, every day to do this like soul crushing, completely unfulfilling, mind numbing job of copy pasting. It's unfulfilling, uncreative and now that person doesn't have to do that anymore.
Yeah, you just save money on the labor cost cost. Right. You also improve the bottom line and it probably, you probably limited mistakes. I'm probably, I'm assuming.
Yeah, well huge issue. Right. If you don't, if you fall behind on notice of commencements or permits, you have to stop work on all jobs until you get caught up on your paperwork. Like you can literally have a stop work order for the entire company because of one piece of paper missing on one job.
Yeah, that's a big, big problem. How would you build that as like you, you probably use n a n8n like and people don't know what n8n is basically just a more advanced zapier and make where you can kind of have a developer kind of create the systems for you. Is that what you probably use that to, to kind of build that, build.
That out the market of. Yeah. So the, the, the software is like the short answer is yes and the market is Opening up to like there's all kinds of new, like there's like one called Active Pieces and Gumroad and there's so many new integration platforms coming out. Agent AI like there's so many of these that exist now. So you could do this with a lot. Because that headless browser is just a node. It's just like a block on that automation thing. It's an HTTP request, which all me, all that means is just like a ping off of a website that says, hey, when you get a ping, go take this action. And then you go to that particular tool and build out what that action is. So you could use a lot of tools. N8N is kind of like the hot one right now. Got a big influencer network around it. So. Yeah, but you could do it in make or even zapier if you wanted to.
Yeah. If you, if you go on, on X, you'll see those screenshots of those workflows that they look super complicated sometimes they're not really that complicated, but they seem complicated that they just look cool. Yeah, they look cool, exactly. And sometimes it's really, if you look at it, it's like the workflows and automations has been around for a long time. It's like that, that really hasn't really changed a lot, but it's kind of like the now the AI twist really. And has it has changed it a little bit. It's more, more advanced and more unique. But that's kind of like, you know, those things are kind of been around for a long time. It's like people don't really understand that.
Yeah, yeah. There's this there meme that goes around and it's like mom, it's like a little boy asking his mom a question. It's like mommy, why are we rich? And it's like, and then mom says because daddy started calling automations agents, honey, you know what I mean? Like now like things that aren't even an agent, people are calling an agent. Like, you know, it's like if it's so, to some of us it's, it's like one of those things where it's like do you join the party or do you kind of like try and call it out? I, I, my opinion on it switches daily. But yeah, yes it is. It's just like in make or Nadan or Zapier, like you have these like nodes, these little blocks. Right?
Yeah.
That connect together and now there's just a new node that's called Agent and that agent requires three pink Things to be connected to it. A large language model. So open AI's models or Gemini's models or Claude's models or X's models. And then you have a memory, short term and long term. And then you have a, a tool that it can use the two exciting things. Well, I don't know how nerdy we want to get on this, but.
Yeah, yeah. Can you walk us to like a typical workflow? I know you talked about the permit one, but is there another one that you built out or you've seen clients built out that can really help improve their operations or their business that you've seen?
Yeah. So a, like the one that like we're really excited about right now is, you know, everybody in. To me, my algorithm makes it feel like everybody and their uncle knows about voice agents. So sometimes I feel like I'm just like another voice in this like echoing stadium of people talking about it.
Yeah.
But when I talk to home services contractors, they're just seeing like an echo. So it's for me it's not as that innovative. But the whole idea is that the ability to make it so that something answers the phone and can be helpful to a customer and then take action in your business is huge. And so essentially what we, what we saw, we tried like 60 different tools out there that were like considered a voice agent. So all that is, is it's the tool that it has access to is a voice. Right. And then the large language model allows it to have a conversation no different than when you chat with it. And then it has a memory short term and like what did the person say two minutes ago? And also has a long term memory. Like what is the context behind my business? Because ChatGPT can tell you what Jeff Bezos said on stage at some thing 20 years ago, but it can tell you what your PTO policy policy is. You have to give it a knowledge base. But then when you do that and you connect it to a tool of voice, that's a voice agent or another tool which is a phone number. Right. Well then now a chatgpt can pick up your phone for you essentially. Now we tested all the tools and the limitation that we found was that when you try and make this thing pick up the phone, it'll just say something like, oh, we can, we're available between 9 and 5, Monday to Friday, when do you want to book? And so in practice deploying this into a business, it was like, oh, well then the agent would like book an appointment on the north side and then book an Appointment on the south side and then book an appointment on the north side. You know, you and me, our businesses, we just go from zoom to zoom to zoom to zoom to zoom with no drive time. But home services contractors have that problem in the case of like roofers. Also if you're in a market where it like rains two days a week and you can't really be climbing around on a metal roof in the rain, well, we have to understand the context of they have a metal roof and it's raining on Tuesday, we can't go to that house on Tuesday. And the other thing is that not all the time do homeowners recognize that their address kind of is very similar to another address that might show up in the same city. Like where I'm from, in a city called Calgary, you have like a 32 62nd Street Northeast and a 32 62nd Street Northwest. And that one character means an hour drive time difference. So it goes and looks at the house on Google Street View takes a picture of it and confirms that with the homeowner. Like you live in a white house with blue trim and a red door. Yes. So we try to find a tool that could do all that and it couldn't. So we had to kind of custom build that. That was, is something I'm really excited about because the numbers behind it is we had a client who was getting 369 digital leads a month. 61 of them, they had never booked an appointment. And when we investigated that, 80% of them were evening and weekend leads. So people who reached out in the evening, everybody who bought a roof would rather bought a boat. So nobody is like, they just want to find the fact like the fastest pathway to meaningfully move towards solving their problem. And if you call them back an hour later or on Monday, it's like someone else who wanted it more has already got the appointment. And this homeowner makes an arbitrary decision whether to pick up the phone, take the call, like, whatever. So this thing being able to meet people where they are, the firefighter who works night and is shopping for a roof while he's at the fire station at 10:30 at night sitting around doing nothing but work out and wait for a bell, you know, he needs to be able to be met where he is. And to put this intelligent thing in place allows us to do that. So then that turned around. First of all, those 369 incoming digital leads, every single one of them requires a 10 minute phone call. So we just gave their Inside Sales Reps 3,690 minutes back for the month, their close rate went from 28% to 30.8%, which translated to another $160,000 in sales. But then those 61 leads like this was pretty consistent for them. They're in the Carolinas. It's a pretty like, steady lead flow. So the next month they booked of those missed appointments. They solved that about 80% of the time, which at that 30, 30.8% close rate, it translated into another. I can't remember the exact number. I'd have to go find it. But it's another. I think it was what, another $100,000 in revenue. I mean, when you do 61 times 0.8, that's going to be 48. 48 deals times 15,000, you're talking. Yeah, like 300,000 sales. Like, it's a huge number. And the other big thing about AI that I have to say in this is that it allows software to move from an access. Like you pay for the software a monthly to be there. Now we can have software, we get to pay for results. So the way that we priced it was instead of paying monthly for this thing to turn on or to pay for minutes of time, we just charge people for the appointments that it books. And this isn't a sales pitch. It's like I'm trying to reframe the way that people on the other end of this podcast think about software and buying things is like, you get to now say, hey, I want to implement AI, but I want to pay per result. I don't want to pay for it to be on. And so the way I look at it is when we, when we sell this agent, it's like you're paying a recruitment fee. Well, we're going to train this agent for your company for a month's payroll of this typical role. Right. Is what about what it costs? And then after that, that employee works 24 hours a day, 365 days a year. It can do. It can take 10 calls at once if you, if it had to, but it, you're only going to pay when it actually books the appointment. So if it books a spam call or calls a customer six times and on the seventh time it books the appointment, you're only paying for that appointment to be booked. So it allows me to kind of feel really good about what I'm selling. Because if I go way back to when I had that school for Facebook ads, what hated was that I sold $400,000 worth of like people who bought the course 1% of people made it to the last lesson. That felt really icky, you know, I mean, because it's like you're paying for access to a course. You bought it. You had the knowledge. Like I could sit and be righteous about it, but it felt icky. You know what I mean? It felt icky that people didn't really get to the end of the course. And really. And I know they didn't take the benefit because some would like get halfway through, like, can I just like hire you? It was like actually the way that we turned into an agent agency. That felt icky.
Yeah, yeah. And how do you, how do you actually build the. You use like the tool. What are the tools you actually use for that? Is that, is that like HubSpot or the CRM? And that's how you keep track of the sales and, and the actual like close rate and stuff like that.
So it'll work with any CRM as long as we can. As long as it has an endpoint for us to create a contact, create an appointment and add an engagement, which most CRMs, like modern CRMs have it. I was surprised your job progress didn't allow us to create an appointment appointment in their system. But it'll work with any CRM that's modern. It just has the three endpoints. And then we use 11 labs for the voice, we use Twilio for the phone numbers, and then we build the language models. It's a mixture of OpenAI and Gemini because we use Gemini when it interacts with Google because it has to interact with like, if they don't have like something like a Rivi or a field service tool, then we got to connect to Google Calendar. It's got to look at Google Maps to calculate the drive times and it's got of a lot look at the either Open Weather or Google's Weather API to pull the weather data. And so we build all of that infrastructure on N8N and then we. But it's like every time it's a little bit different because if they have like.
Yeah, it's a unique, unique for each. Each client, right? I'm assuming.
Yeah, because like somebody will have like rueful and Well, a ruful lead requires it to be aware of the fact that there was a price. And an Angie's List lead requires a level of harassment that you would never put on a bed lead lead. So we have to. And we have to customize them. We also have to build the knowledge base of that roofing company where it can communicate like whether they charge for inspection or not. Or how they handle a realtor call. And so our prompt is about eight pages long, which is really incredible because that agent is like an employee that reads the eight page part of the SOP for your business every time they pick up the phone.
Yeah. And how do you, how do you get the prompt? You get the prompt from the actual. Like, like you work with the actual client from their website, their marketing materials, from the FAQs. Like, how do you, how do you actually develop that?
So the, the eight page prompt is kind of like this is how you handle a call. So you be really akin to like an SOP or like an employee handbook. Like if you had like an SOP for how do we do a lead intake call, what are the questions we ask? What are the variables? Like, that is more the process. The knowledge base is built in a tool called Vectorize. And it's essentially a, like I said earlier, it knows what Jeff Bezos said 20 years ago, but doesn't know your PTO policy. So something called a vector database. And it's just a fancy way of saying it's just a. It's just like a large table of data that allows, that allows us to give the agent a brain that basically says when people ask a question, don't go to like the OpenAI brain, right? It's called retrieval augmentation generation. Retrieval augmentation generation is rag, right? Rag database. And all it's basically saying is retrieve your answer from this augmentation of your brain, which is this database that has our company's knowledge. And so there's a process of like dumping a website, their PDFs, their anything about their company, we dump it into this like intake. And then they do what's called an embeddings, which is just saying, okay, go put all this data into a way that like a large language model knows how to access it as a brain. It's simpler than it sounds. But. But that puts this knowledge base together that allows that agent to have a brain that's unique to the company.
Yeah. I mean, thinking about what you just said, this really applies to like most trades like the H Vac Gazette and Plumbing Plus Control, they all have to drive to get to an appointment, right? It's dependent on the weather a lot of times with these trades, right? Depending on the season, right. Depending on the zip codes, all these different things that this actually, I mean, all these trades that have like that are driven by the actual weather by the time of the year. This is something that they can build themselves, right? It's something and apply for them. It just has to be like customized to each, each individual business and, and, and trade. Right, that's, that's great.
You get to customize the voice.
Yeah.
Right. Like some people have like a British guy, other people have like a, a young girl. Some people have Bart Reynolds for a voice.
Yeah. What do you think some of the fears that contractors have with, with AI? How do you think they should overcome them?
Well, there, there's a general fear that it's going to cause us to like lose jobs. Right. So like when we think about the smaller contractors, they're, they're thinking like, well these big companies are going to come in and do all this AI stuff and then they're going to become so much more financially effective. Right. And there's this fear, but it's actually if you just. Those companies are really hard to pivot. They're really hard to like move and maneuver and they just get a lot in their own way. And then the smaller companies, they're really in a place right now where they don't have the bureaucracy of a bunch of people. One of like the greatest adopters I've seen is a guy from a company called Action Roofing out of Canada. He is like lean or another guy dressed Drake who sells Christmas lights ironically also out of Canada because they shut down half the year, you know what I mean? Because of the weather.
Yeah.
So they're forced to like lay off people every year and rehire people every year. So the idea of having a dormant ambient AI that just like doesn't require a layoff and doesn't require like the emotions of retraining, rehiring, recruiting, they're really, they did it. And like Action went from a 20 year long, long business that did like one and a half million a year and then he went to, to 4 million in a year. And he's just embracing AI at every step of the way because he's just thinking about that such lends to this fear that it's going to create a lack of jobs or some kind of like job decimation. But every example of this in the past, like internal combustion engines, the Internet, it's always created more jobs. It's always created more jobs. Because a great example is if you haven't heard, if you, if you haven't watched the news lately, maybe you've heard this. But like these AI companies are building their own nuclear power plants. They're building huge data centers. This is creating so much work in H Vac. Like you know how Much H Vac. You need to cool a data center. Like they want to build them in Canada because it's got natural air conditioning, right? Like the amount of plumbing, you know what the, you know, like have you looked at what the size of like a, like that supercomputer that runs Grok, how big that building is?
It's massive. Yeah.
A massive roof. And do you know what the cost of a leak is on if you have of water drip down on a tower that has 20 GPU 1000s in it, that's like a two and a half million dollar bucket of water. Like the amount of work and, and the expectation of high quality construction that's going to happen as a result of this, you're going to see the greatest boom in the trades business that you've ever seen.
I agree with you. I agree with you. I think, I think there is, there will, there will be some job loss in the white collar section. Actually a lot of white collar jobs will, will be gone I think depending on what kind of jobs. But a lot of them.
Yeah, but you, but you could have like a RO company is going to have less need for a permit person but they're going to need an AI engineer. So I'll give you an example from myself personally. We have a person here named Priscilla and she's an HR person, right. We went through, we had an employee basically do a horrible thing to us left and stole $100,000 in monthly recurring revenue. We had to lay off 30% of our people overnight because they, they took two employees and $100,000 worth of clients. And I get why the clients left they that we edified this person for seven months to them and when that person like and we can go to court and that's all happening. But that doesn't stop the hundred thousand dollar loss when they were like well we just want to follow our guy because we don't want to hurt our business so we're going to follow this guy. I get why they made the decision. I have empathy in my heart for that. But doesn't change the fact it was hurt. Well all of a sudden HR had 30, 40% less work to do because when you lay off 30%, another 10, 20% say ah, I'm going to leave too. You know what I mean? Like, like, you know it creates people brushing up on their resume. Well that HR person now suddenly had less work. Do we let her go? Well, she took it upon herself to go into like three different AI school groups. She's watching everything. She can Learn about AI. She's implementing hi in the HR thing, but now she's kind of moved into being the executive assistant for the director of operations here. And she's now building a. Building a private app in the QuickBooks developer workspace. And she's trying to build some kind of like automated CFO type thing. This was not her career path. She just, she was just suddenly allowed to be creative and to be a problem solver and I just didn't get in the way. She just, she just went and pursued this on her own and I just supported it. I didn't say, no, this is your job. Do your job.
You know what I mean? Yeah.
And now she's doing like six things that are completely different. And it's amazing. Yeah, I think that that is what you're gonna see. You're just seeing this. You're seeing a decrease in need. Need for certain things that can be automated, but then a massive need. And it's like you literally just can ask ChatGPT to teach you how to do it. You don't have to go to school. You just be like, hey, Chat GPT. My job's being eliminated by AI. These are the skills I have. These are the problems I'd like to see solved in the world. Teach me how to make those problems better. And it will, like, it will literally just. You just keep asking it questions until you get the answer. Like, yeah, I think it's gonna. It has. It is both the tool that creates the opportunity and it is also the function of change that's creating the presence of a need for work. So I think that it's creating way more jobs than it's losing. It's just changing them. It's just changing it.
I could tell, man, you're very, you're very passionate about this. I think I'm very passionate too. That's why I started this podcast because I've been. The last two years, I've been really digging deep into AI, learning it myself. At one point I was spending like two or three hours just going on YouTube, going different courses, just learning it myself and using the tools. Tools. Now it's just basically empowered all parts of my business. And now the big thing for me that I use it for and I hired. Now I have two people that are working on that full time, like automation person and AI person. But for me, I've. I use it to just to think, really. That's the biggest thing, the biggest tool that I use. I use it to think easy, to run better meetings, to write Better scripts, just marketing, obviously, marketing. All the scripts, all the, all the technology that we use it for. But that's the biggest thing, how to, how to use it to think better.
Better.
Right. That's. I think I feel like a much.
Smarter 40 IQ points.
Exactly. Like just. He becomes smarter on it. He becomes smarter on it.
Yeah, there's, there's. There's a study that I heard referenced, right. And it said that if you put a really smart person in a group, right, the whole group operates at a higher residence, like, because they're all borrowing that 40 extra IQ points, so they move through solving problems faster. Like, the proximity to a really intelligent, intelligent person allows a group of people to all solve a problem faster. Well, that has just scaled out by the order of like a billion in that now every one of us has access to that additional 40 IQ points on demand whenever we're inspired to do it, to use it. So just like you said, you can use it to think better because you can process your decision making better. Yeah, it's incredibly empowering for people who previously, the limitation was like, I grew up in Sheboygan and I never got got exposed to this, and now you can just like, for 20 bucks a month, go and be like, open my eyes. Chatgpt. Tell me what I don't know.
Exactly. And there's a couple things. Basically, I think if you work with a team, right, especially my team, I think every, every business deals with this is that there is a kind of a fear with employees using AI, right? Like I, I talk about it myself. Some people in the company are using it themselves, but not really everybody is getting it to, to actually use it. Right. And that's kind of like a typical fear that a lot of employees have. So have you. Have you had. Had that experience before where you. Let's start using AI. You have, you know, 10 to 15 people or 20 people working for you. How do you get them to actually use AI? Have you, have you had that before experience?
Yeah, we, we. When the Shopify CEO sent out the letter to all his employees and said that, like, using AI is now like use basic, the summary of it in the headlines was use AI or you're fired. That was what basically what Shopify's memo said. And when he said that, I had the same meeting with my team and I had a picture of like, Friday the 13th, Jason Voorhees on based, basically saying, like, this thing is coming for all of us. And I did have some resistance. I had a graphic designer who was like, you know, I went to, I went to design school for X years. Like AI can't do my job. And it's like, you know, and so we had some initial resistance and then we had some people who, you're like, hey, here's an axe. Chop down this tree. And they instead try and like invent a chainsaw because they just want to play with AI. So there's two sides, right? People who refuse and the people who, who can't just go do the job. They got to try and figure out a way to get AI to do it for them. And they'll spend six hours trying to solve an hour long problem because they just want to play with AI. So you have both things you have to do. I haven't, I don't think I've nailed it yet. We definitely, I would say have a probably higher degree of AI because it's just another tool, right? The tool doesn't matter as much as the process and the process doesn't matter as much adoption. I've been saying that for years. So it's just another tool. But I think that if you create environment of, if you create an environment of safety where people are free to try things and people are free to make mistakes and have the AI accidentally say the dumb thing or, you know, maybe something comes out with a spelling error because now they're moving faster and they're not catching that the AI generated image has a spelling error. If you create the freedom for that to be okay, then people will try generally to embrace it. And you just have to like, I think give them the assurance that this isn't me trying to get you to do five times the work for the same amount and then fire you. It's like, no, I, I want you to be able to do five times the work so I can pay you twice as much and I don't have to hire the next person.
Exactly, exactly. So let's talk about marketing. What's the number one marketing KPI or metric that any business owner like in the trades and a roofer should, should look for?
Yeah, I actually heard, learned this from a client who is very successful. He sold a health, like a, what do you call it? Like health insurance for seniors business. Like they phoned home seniors and sold them like health insurance insurance or some kind of medical plan. So his whole thing was like, how do I get to the point of pushing a button and getting a sale? That was how he treated that previous business. He sold that for 80 million bucks. But his best friend from high school was an OC Rep. So his friend from high school was like, hey, I heard you're rich, you should start a roofing company. So he started one and he went from like 0 to 0 to 16 million in, in, in 18 months and then 16 to 32 and then 32 to around 42. And they're kind of like now going through another growth. But he said to me when we were building his HubSpot, he said, Adam, I just want to get a lead for 300 that close at 30% at a 40% gross profit, that I can get the money back in 30 days because then it's customer funded acquisition. I just have. And I said, okay, well then how do you quantify that $300 lead? So that's like his top level KPI is 330%, 40% profit, 30 days. Right. And then he said, I don't care what the channel is, if it's carrier pigeon, smoke signals, Facebook, TikTok, Google phoning, I don't now door knocking. I don't care. He looks at door knocking is like okay, how? Like he looks at door knocking as marketing, not sales. And then the next level KPI was I need 15 to $50 cost per thousand impressions. I need 1 to 3% click through rate and I need 3 to 10% conversion rate. Those numbers are going to fluctuate anywhere in between. And then so if I get a hundred dollar lead and when I guess we should clarify $300 lead appointment or marketing qualified lead. So if he gets a hundred dollar lead, but it takes but only one in three of them book an appointment, well then that's a 300 appointment.
Yeah.
Or if you get 30 leads at $10 a lead and only one of those books an appointment, then still a 300 appointment. So like Facebook was a terrible channel for them. They would get like tons of leads, but they would never get a hold of them. Will they put an AI agent in place to contact those leads to decrease the burden on the humans. And then they were able to get it to the point where all those crappy messenger leads did turn into a $300 appointment. And yeah, there's 26 of them that never responded. Who cares? You got a $300 lead and your sales process closes them at 30% at a 40% gross profit. And if you can turn that into cash in 30 days, you're getting paid to service the first lead before you have to pay for it. So you can afford to buy the second lead in the same period of time.
Yeah, that's amazing. So he came up with that number, I'm sure like, like through reverse. Like he, he basically okay, this is what the cost is actually to actually sell a roof. Right. And I need to figure out how to make a profit based on how much money I'm going to be spending on the actual qualified lead. Right. So it's. He works I'm sure partly backwards, right, where he actually.
Yeah, you work your way back because the average ticket at the time was like 18,000. Right. So now if the price of the roof has increased to 21,000. Right. Well then maybe the cost per lead creeps up to 450 because his cost of acquisition. If you take 450 and divide it by go 450 divided by 0.3, that's 1500. 1500 divided by 20. I would say 21,000 is 7.1%. So you know, he's paying. A lot of people want to spend 5% on marketing and that's what they say. I want to spend 5% on marketing. It's like that's great. Maybe you're spending 5% of your current revenue, but maybe you got to spend 7% of your current revenue to get that to eventually be more money. Right?
Yeah.
So he forward investing for market share because he's looking at it as a land grab. Right. So he spent a little more money on marketing than other people but he's not afraid of a small business. He sold his last company for $80 million. So like that's a multiple of profit. He's not even at $80 million in revenue yet. So he's investing to build something big. I don't expect everybody to think like that. It's just that you can.
Yeah. What do you think the. This is a very common argument. I've hear like all different like things opinions on this. But what do you think the average contractor should be spending on marketing budget wise per year?
I think that the numbers evolving right now you have private equity trying to do a massive land grab because they all rushed in and they spent like the short answer is like 5 to 10%. But like that's a huge difference. Right. And it depends on the company. But like we have to think of like what are the. The three. There's three tectonic shifts, like massive plates moving underneath our feet. One of them is capital constraints. Like interest rates are stuck stagnant, stagnant growth. We have all this like issues with the economy and capital and everything like that. So like there's just the capital markets are constrained. Second one is in. In My industry, roofing specifically. I don't know what it looks like in like plumbing or I know plumbing, H Vac, see this. But I don't know, like arborists, you know, but private equity is coming in and their whole mode is to buy a company for 2x, smash a bunch of them together and then sell them for 5x or 3 and 7 or 5 and 13. Right. They're just trying to create value and then package all up and sell it. And then like they do this over and over again until it eventually becomes like Ford, Dodge and Chevy of roofing. And they're on a public market. So that is creating this like thirst for growth at any cost. And so they're like, like slashing costs everywhere to try and pump more money into marketing because they're just trying to win the race of that next exit before another company does. So they're willing to like burn cash on marketing because it for them, that's value creation.
Yeah.
So you're competing in that market. And then the third tectonic shift is AI. Right. Is like orders of efficiency. So see, private equity has money. So they don't have a capital constraint. They have money. They don't have to go get it. They have have it. Some of them use leverage, but for the most part they have access to capital like we don't. Then they also have a different incentive. Right. They're getting a multiple on the sale. They're not necessarily worried about being around for 30 years and building a legacy. They just have multiple expansion. And then AI is something that they can invest in to say, well, if I can bring this cost down 2%, I can immediately take that money and put it into my marketing budget. And then you have to compete in that market either by being more effective or just spending more money. And so you either have to get more back bang for your buck or you got to spend more bucks. So you have to decide which one of those two things are you are. And that might mean that you know, the guy who's willing to get on Facebook live every single day, twice a day can get more bang for a buck. They can spend five but be more effective. And then yeah, equity will just bury a market in cold callers and, and flyers. But they're going to spend seven.
Yeah, exactly. Yeah, it's, it's a tough thing because what I've seen is that I've seen, it's the rising, the cost is rising every, everywhere. Right. And on ads, on the, it gets more expensive to acquire a customer in Google, on Facebook These channels, there's more compete, more competition, there's more people doing lsa. Like it's, it's competitive everywhere.
Right.
So it's more getting more and more competitive. So it's, it's tough because a lot of contractors are hesitant to spend the required amount sometimes to be successful. I think based on my research, I think the really just. They say you say 5 to 10%. I say very few contractors are actually spending that amount. I would say probably the average is 1%.
Oh yeah, they're not doing it.
Yeah.
But like we did a secret shop of roofing companies. We built an agent that sent to 2,000 contractors. That basically that headless browser thing. It logged on to 2,000 roofing company websites at 7 o' clock at night on a Friday and it filled out the contact form, sent the lead and we used a fake email and a fake phone number and we monitored the responses. And so it says like what, who was it? It's like New Orleans OR no, Louisville, Kentucky. Right. There are 61 contractors there. Bone Dry Roofing was like the runaway winner on lead response. Instant response, instant human call, instant email, multiple attempts to follow up. Very perseverant. Lots of like really good, engaging emails. Then there was like three other companies that were kind of like in the top and then it was like 50ft of crap below that. So Bone Dry Roofing doesn't need to find a new marketing agency. They know what they're doing and they don't need me to come and build them a system or an automation. They know what they're doing. But the problem for the 50ft of crap below that to say that they just have a bad lead response or bad marketing is too, is too surface level. And I like to relate it to things that people can understand because they, because they're the person that if you or me or any marketer cold called them, they'd be like, oh, you guys call me all the time. I get all these calls. Everybody, everybody promises they can do that. Like they like they're crusty to sell marketing to. Well, if you saw for like most of these contracts, directors, when you look at them online, you can see that they're doing a shitty job like you or me or anybody in the space, like in the service side of this industry, we can see what they're doing and we can see that they're doing it wrong. If you saw twelve 400 pound people sitting on the side of a sidewalk and they were like £400 just standing there eating ice cream, if you're in the personal training health space, health food mindset coaching business. You would eventually figure out, hey, on Brown street, there's like twelve 400 pound people standing there eating ice cream. And every day somebody woke up, hey, I can teach you to eat better. Hey, I can teach you how to work out. Hey, I can teach you how to get your mind right so you can see and believe in yourself to lose weight. Hey, I can teach you how to walk 10,000 steps a day. Have you tried buying a step counter? Have you tried buying a walking treadmill? Have you tried like this or that? I could help you find groceries. I'll cook for you. I'll do a meal planning service. After a week they'd be like, oh, there's all these people coming all the time. And like I tried it a few times, like, and it's like, great. Did you commit? Like, how long did you do it? Like, well, I walked for 30 days and I didn't have a six pack. Okay, well, it takes longer than that. Well, they told me it would be. It's like, that's because they're trying to sell you, but they would get crusty to deal with. But it's like the reason that you're getting these calls is because it's obvious that you're not solving the problem because everybody can see it. And so you still need to be sold the solution because anybody who's in the business of selling the solution knows where to find you, what to look for, and you display the trick traits of the problem. And so, yeah, it's like nobody's calling bone dry to sell them an AI agent. Nobody's calling bone dry to sell them marketing because they already got it figured out. Yeah.
And they're one of the biggest guys out there that I've read. I mean it's called, I call it curb appeal. Right. I think contractors instantly understand that if you show up in a rusty old truck, right. It's gonna hurt their sale process. Right. It's gonna hurt their possible sale. But a lot of times if you have a crusty old website that looks ugly, it's what, that's what, that's what, that's a perception that con that these customers are having, having when they see you online. It's like an old, rusty old truck. It's not, it's not gonna. Not trustworthy.
Right.
It's, that's why it hurts contractors first. Right. So I really think it's, it's a, it's also a big educational thing too as well. A big educational Thing. It's a mindset issue as well. And at the end of the day there is a lot of marketing companies out there that over promise and under deliver. That's a big part of it as well. But what. Okay, so besides those, like can you talk about some of the like the biggest. I want to hear this. Like what is the, the most successful campaign that you ever ran for yourself or your, for your. Or for like a company or a client that you, that you had like a marketing campaign?
Yeah, the best one I've ever run was a 89 gutter cleaning ad. This business, it was, we took the lesson. They had a door knocker who to knock on doors. He was saying hey, I'll clean your gutters for 99 bucks. And he got on roofs a lot more. When he offered to clean the gutters for 99 bucks, he knock 100 doors and say do you want a free inspection? You get one or two. He knocks on door, says hey, I'll clean your gutters for 99 bucks. He'd get 10. Then when he was up there, he'd be like, hey, by the way, did you know you have damage? Want to file a claim? So let's now apply leverage to that. Run that as an ad. $99 gutter cleaning. So then, now we just created the gutter cleaning service. And I mean I learned this lesson like 6 years ago too because the, the, what do you call it? The a guy had his dad who was an older fella and just gave him a job doing gutter cleaning and his dad would be like, well you know, you're looking kind of bad and talk to my son. Right? So I had two examples of this in real life and it was. And you all hear about like Alex or Mosey and, and lead magnets and stuff like that. It's like what is a lead magnet? It's either a partial solution to a problem, a temporary solution to a problem, or the sudden awareness of a problem. And it's usually where you give a high amount of value for a low cost to get the lead. Well that gutter cleaning thing, it just got the same customer where you got customer funded acquisition. It costs you 50 bucks for the lead. But you were paying, getting paid $99. So you're paying to get the lead. So you could dump as much money. Even other campaigns that maybe go got more leads or like, or, or, or we're like more like roofing, I want to roof leads. But the thing was is that like kind of ping pong to two step approach Created a profitable lead generation.
Got it. And then you basically ran it to a landing page and did you have like a checkout form on there, like where people can actually pay online or.
Yeah.
Okay.
So it meant that we had to like, convince certain salespeople that to get those leads, they had to be willing to go and like spray down a gutter and on, unplug a drain. You know what I mean? It wasn't a hard job to do. And then you charged more money if there was more gutters. Like, it just, it just. You had to create a company that was willing to do that, but then you created a marketing department that was profitable. Like, the salespeople made more money because they didn't have to go knock doors. Now they just scaled that up and then. But they had. Instead of spending five hours knocking doors to get two doors to let them on a roof, they spent an extra 35 minutes on a house cleaning. Cleaning gutters, you know what I mean? But now they didn't have to knock anymore. So it was a change in just effort. What effort are you putting in to create the opportunity for your sale? Is it walking or is it cleaning?
Yeah. Awesome. Awesome. So, yeah, where can people find you? Where should people look you up?
You know, I found out recently that more of my, like, my best customers all seem to like X. And I don't go on X a whole lot other than like read the news, you know, I mean, like, I don't participate and post much, so I'm like, this is like the first time asking this, but like, follow me on X at roofing partner or you can just go to roofingbusinesspartner.com but like, I'm just like, I want to find some friends on X. And I found that like, I was like, wait a minute. More of my customers are on Facebook, but more of my best customers are on X. So if people are on X, I'd love to connect.
Are you going to start posting content there?
Yeah, I think I'm going to start like, just. I like to write and so I think. I think I'm. I want to participate more there. So I'm trying to of like, give myself some dopamine hits for being on there.
Yeah, I think I. I've started writing on Facebook again, but I think X, I want to start doing it too as well. It's a. It's a good platform. I do see a lot of successful contractors already on it, surprisingly. But I mean, I think Facebook has been the number one where people find each other. I think. But I think X is the next big one. But. But Adam and it was awesome. Thank you for everything. I really appreciate it.
Of course man. It was awesome to be here. Thank you and have an awesome day.
Thanks man. Have a good one. Bye.
Thanks for listening to AI and Marketing for Home Service Pros. If you got value from today's episode, do us a favor. Follow the show, share it with a contractor friend and leave a quick review. Want help using AI or smarter marketing to grow your business? Head to contractormarketingpros.net and book your free strategy. Call.










